Testing Seeds for Purity and Origin

ALBINA F. MUSIL.

THE SCIENTIST who is called on to analyze seeds may have to determine the purity of a sample of seeds, examine it for seeds of noxious weeds, find out its origin, and ascertain the varietal purity of the seed stock.

The person who submits the seeds to the laboratory for test usually specifies the services he desires.

The purity test determines what proportion of the sample is pure crop seed and what proportion is seeds of weeds, seeds of other crops, and inert matter.

Assuming that the sample he receives is of adequate size and is representative of the seed lot, the seed analyst or technician reduces the bulk sample he receives to a smaller sample on which the test is to be made, usually referred to as the working sample.

Most kinds of seeds are mixed and divided by means of a mechanical divider, a device that mixes and divides the sample into two parts. The dividing process is repeated until the desired size of sample is obtained. Samples of seeds that are not free flowing, such as cotton and certain chaffy grasses, must be divided by hand. The seed is poured into a pile on the table, mixed thoroughly, and halved repeatedly until the amount is right.

The minimum weights to be used for working samples of agricultural seeds, vegetable seeds, and certain herb seeds are prescribed in the Rules for Testing Seeds, formulated by the Association of Official Seed Analysts, and the rules and regulations under the Federal Seed Act.

The amount varies with the size and nature of the seed to be tested. Seeds as small as redtop (Agrostis alba), which averages about 11 thousand seeds per gram, require 0.5 of a gram; Kentucky bluegrass (Poa pratensis), averaging about 4,800 seeds per gram, requires 1 gram; hairy vetch (Vicia villosa), with approximately 36 seeds per gram, requires 100 grams; field pea (Pistim satimm), with only about 4 seeds per gram, requires 500 grams.

To avoid errors due to any hand manipulation, the bulk sample is reduced to the nearest minimum weight specified in the rules. The working sample may exceed the prescribed weight but should never be less.

After the bulk sample has been reduced to the prescribed size, the working sample is weighed and taken to a workboard, where it is examined for trueness to name. The sample is then separated into its four components: Pure seed, other crop seeds, weed seeds, and inert matter.

Because of the small size of the units that make up most seed samples, the separation must usually be made under magnification. A hand lens of 6 X or 7 X magnification is suitable for most samples. The lower magnification of an ordinary reading glass is adequate for large seeds. The technician must be able to identify correctly every particle in the sample.

The component parts are weighed on an analytical balance that can weigh accurately to three or four decimal places. The less sensitive torsion balance may be used for large seeds, such as beans or peas. The percentage of each component is calculated on the total combined weight, and the kind and number of other crop seeds and weed seeds are recorded.

As a check against possible loss of material in the process of making the separation, the combined weights of the components and the original weight of the sample are compared. Another test is made if there is an appreciable difference between the two weights.

The evaluation of damaged or undeveloped seeds may present difficulties. Seeds of crops and weeds may be broken or otherwise injured, or the seed unit may be of such nature that it is difficult to determine visually whether a grain or embryo is present. The technician must determine whether they should be classified as "good" seeds or as inert matter. The standards for evaluating crop and weed seeds of this nature differ in some important respects. To assure uniformity in interpretation among seed technicians, the rules give specific instructions for the classification of questionable units.

A damaged crop seed, such as a broken alfalfa seed, obviously should not be counted as two seeds instead of one. In such cases, the rules provide that pieces of crop seeds larger than one-half the original size are to be classified as pure seed; pieces one-half the original size or less are classified as inert matter.

The broken part that is classified as pure seed may or may not produce a plant. Its planting value will be determined later in the germination test. The determination of potential germinability of agricultural seeds is not considered a function of the purity test.

The rules provide that diseased agricultural seeds shall be classified as pure seed. Those whose contents are replaced by fungus bodies such as ergot or other sclerotic, smut balls, or nematode galls however, are classified as inert matter.

Damaged weed seeds require a somewhat different interpretation. Because weed seeds are not tested for germination, as agricultural seeds are, the seed sample is not penalized by classifying as "good" weed seeds the seeds that are too damaged or undeveloped to grow.

The rules specify the conditions under which a weed-seed unit shall be classified as inert matter.

The following examples illustrate a few cases of damaged weed seeds or seedlike structures that would be classified as inert matter: Hulled dock (Rumex), more than half of whose embryo is missing; seeds of dodder (Cuscuta) that contain no embryo; bulblets of wild onion (Album) that show damage at the basal end or are devoid of husk and pass through a 1/13 -inch round-hole sieve; immature florets of quackgrass (Agropyron repens), in which the caryopsis (grain) is less than one-third the length of the palea; empty hulls of seeds (achenes), such as the sunflower (Helianthus) or the docks (Rumex).

The absence of embryo or endosperm may have to be determined by dissection or by examination over a diaphanoscope. A diaphanoscope is a device in which a strong beam of light is directed upward against a pane of clear glass, over which the seeds are examined with a hand lens. A light intensity of about 200 foot-candles is usually strong enough to penetrate the hull or seedcoat of most kinds of seeds.

Special techniques sometimes are used to establish trueness to name of the sample.

We do not attempt to distinguish individual seeds of perennial ryegrass (Lolium perenne) and Italian ryegrass (L. multiflorum). The possible diagnostic structures, such as awns and pubescence, usually are removed or damaged in harvesting and cleaning to such an extent that the seeds are indistinguishable. The fluorescence test is used exclusively to detect seeds of Italian ryegrass and hybrids in lots of perennial ryegrass or to determine perennial ryegrass in Italian ryegrass.

The fluorescence test involves a germination test. Four hundred seeds are taken from the pure seed fraction of the purity test and germinated on filter paper. The seedlings are examined under ultraviolet light, and the percentage of fluorescence and nonfluorescence is determined.

The results are then subjected to either of two formulas to calculate the proportion of each kind in the sample. The first is: The percentage of perennial ryegrass is 1.0526 times the percentage of nonfluorescence times the percentage of pure ryegrass divided by the percentage of germination.

The second formula is: The percentage of pure ryegrass or hybrids, or both, equals the percentage of pure ryegrass times the percentage of fluorescence minus the sum of 0.0526 times the percentage of nonfluorescence divided by the percentage of germination.

The formulas take into consideration a small percentage of short-lived perennial plants that may normally be present in perennial ryegrass.

A workboard, which elevates the working area above the level of the tabletop and so minimizes fatigue of the eyes, neck, and shoulders of the technician who makes an analysis of the purity of seed.

We have no way to tell individual seeds of white-blossom sweetclover (Meldotus alba) from yellow-blossom sweetclover (M. officinalis). Yellow-blossom sweetclover produces a variable proportion of seeds that are spotted often faintly with purple. They are referred to as mottled seeds. As far as we know, the varieties of white-blossom sweetclover grown in the United States do not produce mottled seeds. When mottled seeds are observed in a sample labeled "white-blossom sweetclover," it can be assumed that seeds of yellow-blossom sweetclover are present. Most samples from plantings in fields and greenhouses have indicated that about four times as many yellow-blossom plants will be produced as there are mottled seeds. We conclude therefrom that each 1 percent of mottled seed represents 4 percent of yellow-blossom sweetclover.

If a sample is submitted as white-blossom sweetclover, a mottled-seed test is made to determine the possible presence of yellow-blossom sweetclover. A minimum of 400 seeds, taken from the pure seed fraction of the purity test, is examined. The entire 5-gram working sample is examined in some laboratories. A hand lens or a higher magnification may be necessary to detect the fainter markings.

The rules provide the following formula for the determination of percentage of yellow-blossom sweetclover in a sample of white-blossom sweetclover: Weight of mottled seeds (in grams) times 4 times percent of pure sweetclover.

Several new varieties of yellow-blossom sweetclover have been introduced. Some of them show great variation in content of mottled seed. There is strong evidence that the area and year of production strongly influence the percentage of mottled seed. These findings indicate that our formula for determining the extent of an admixture should be considered only an estimate rather than an accurate analysis. A field or greenhouse test should be made when an accurate determination is required.

PELLETED, or coated, seed may contain single seeds or several seeds each. The pellets are designed to make small or irregular seed units easier to plant. The coating material may be inert material or it may contain some fertilizer or fungicide.